lots of bad sectors, disc restarts pc, now what?

[wjgollatz]

I have had a series of problems with my single hard drive.

At some point, problems caused the OS or BIOS to have the SATA ports operate in PIO mode instead of DMA. I got that resolved. But system restarts after long disc operations is now occuring (primarily when error check and backing up).

I used HD Tune to measure read speed, and it was much lower than it should be, but higher than the PIO mode. It would start great, then dramatically drop to 10 mb/s, Then jump up to 100 mb/s, then back down to 10 mb/s for a bit, then back up, and oscillated at highs and lows for an average of 30 mb/s.

Using HD Tune to check for errors, I soon found lots of consecutive red/bad sectors, then the system restarted again afer what seems to be a short freeze when HD Tune was showing the most recent sector as bad. It seems that operations on the bad sectors slows the disc down, then causes the pc to restart (after a short freeze).

So - what can I do. I wasn't able to even complete a error check. In fact, I have been unable to even do a backup of the C:/ partition of the drive (just 4 total partions on a 500 gb drive), because after some time passes, the system freezes shortly, then restarts. I suspect it operating on bad sectors. I believe these bad sectors are on the C partition. HD Tune finds them relatively quickly, and when I did the partitions, I had the C partition on the outside of the drive for the faster rate of being on the outside edge as opposed to the inside edge, and I assume HD Tune starts on the outside edge because the access times are always faster at the beginning of a test.

So - what can I do? Is there anything I can do? What is the best way to retrieve data?

 

[VirtualLarry]

Use Norton Ghost 2003 for DOS, with the "-fro" switch?

 

[Cerb]

I typically use dd, from a live CD, to an equal or larger sized HDD.

dd if=/dev/baddrive of=/dev/gooddrive conv=sync,noerror

sync pads unreadable blocks in the output file (may not be needed if not setting bs, necessary with large block sizes), noerror keeps going, instead of stopping, on and error.

Once copied, make a copy of the copy, so you can indiscriminately use testdisk and chkdsk on the partitions, rewriting the backup image if they screw data up, instead of recover it.

I've also had very good results using Linux' NTFS for file recovery of known corrupted partitions, and known failing drives. The drive must be mounted from a root terminal, as every automatic GUI tries to use NTFS-3g (which is better for a healthy partition).

The last few years, I've had poor luck with old bootable discs and such, like the old Ghost, even with everything set to IDE (also UBCD, HBCD, and Samsung's and Hitachi's HDD diag ISOs). I'm guessing that either mobo makers, or ODD makers, aren't testing for Dr. DOS compatibility much, these days. I haven't done any rigorous testing to see if the problem is mobo or ODD firmwares.

 

[razel]

The best way to retrieve your data is to take your drive out and use it in another PC. Once you have your most favorite things copied it's up to you whether to try and debug what happened or just start from stratch.

Either way, those auto restarts are what I'd worry about. They could be causing the disk errors. What'd I do? When you are on that other PC, create a boot disk for memtest+ and run it on your computer. You just need one pass. While that runs download your drive manufacturer's diagnosis utility and test your drive. You just need to run the SMART and short test. You don't need the long test for now.

Best of luck.

 

[wjgollatz]

I have found that my pc runs fine, until I start copying certain particular files. Then the pc will slow down, freeze and restart. Perhaps the restarting has soemthing to do with the disc errors. Right now, I am using the PC in question and if I try to copy a particular file, I can make the pc restart.

I wish I could identiful what files were corrupted, set them aside, and copy everything else off.

All my *important* files are copied. But lots of stuff I'd still like to keep.

 

[Cerb]

You need to get any data you may care about off of there, booting from a CD or another HDD, and go from there. When you either get it off, or determine you can't, replace it. If you want any more data off of it, you should only be powering it up for the purpose of data recovery. If downtime is important, pay a few bucks extra for the 1-year warranty drives at BB.

 

[taltamir]

You need to get any data you may care about off of there, booting from a CD or another HDD, and go from there. When you either get it off, or determine you can't, replace it. If you want any more data off of it, you should only be powering it up for the purpose of data recovery. If downtime is important, pay a few bucks extra for the 1-year warranty drives at BB.

that, furthermore.

If it has important data, stop booting from it. Windows loading off of it and modifying files as it runs is not doing you any favors in terms of having a chance to recover your data.

Take it out of the PC, fix your PC with a new blank drive. ONLY THEN connect this drive as a secondary and copy what you want while you still can.

 

[iloveATech]

If it has important data, stop booting from it. Windows loading off of it and modifying files as it runs is not doing you any favors in terms of having a chance to recover your data.

Take it out of the PC, fix your PC with a new blank drive. ONLY THEN connect this drive as a secondary and copy what you want while you still can.

!!!!!

Listen to this!

 

[sm625]

Bad sectors tend to multiply in an unpredictable manner. I dont even attempt to work around them. Bad sectors = refrigerator magnets.

 

[Binky]

Have you considered that the cable might be bad?

Take the advice above and stop booting from it until you pull important data from it.

 

[jjsbasmt]

Others have given you great advice, but I'm just a bit curious about the drive. Brand, size, PATA/SATA, age and number of hours on the drive and normal operating temps of the drive. I've replaced scores of bad HDs over the years for people and I'm always interested in the particulars of the drive. Furthering my KB. Thanks

 

[wjgollatz]

Thanks for all the advice. Because I am getting a new HD I haven't been using the drive much becaus eof the bad sector, waiting for a cheaper drive to come buy - I hadn't relized that drive prices were still so high.

All my important data is off, but there is other stuff I'd like.

 

[wjgollatz]

Others have given you great advice, but I'm just a bit curious about the drive. Brand, size, PATA/SATA, age and number of hours on the drive and normal operating temps of the drive. I've replaced scores of bad HDs over the years for people and I'm always interested in the particulars of the drive. Furthering my KB. Thanks

Seagate Barracuda 7200.12 ST3500418AS 500GB 7200 RPM 16MB Cache SATA 3.0Gb/s 3.5"
http://www.newegg.com/Product/Produc...82E16822148395

HD Tune gets this data Power On Hours Count: 14,110. Which doesn't seem right from a 2/28/09 purchase. Normal operating temp is 89 or less. Been 89 since I have been examining it - but that is with a wide oen Antec 300 case, and no good airflow.

Its definitely bad sectors in areas of the drive, I rarely ever use. I was only goign through data so see if I wanted to keep it, and noticed copying problems and slow speeds. The case is situated that it is never bumped and is not even top heavy (PSU is at bottom), and case has 2 large fans. I have no idea how long its been bad, becAUSE EVEN NOW, all OF MY PRIMARY APPLICATIONS WORK FLAWLESSLY (for now).

For your killboard, I only first noticed a problem when all drive speeds on all ports went to 1.8 mb/s. Later realized that Windows does something if there is so many conflicts or something and the ports were reduced to PIO mode instead of DMA. I got that fixed, launched HD Tune again, and the read write speeds averaged at 30 mb/s, with LARGE frequent oscillations in speed from 50 mb/s to 6 mb/s across the test. Normally the drive was at 125 mb/s to 75 mb/s. After that started doing quick backups before I did a full disk backup and then noticed copying problems, anf then Acronis won't even back it up and neither will a Acronis boot disc clone the drive. My bad, I got last and was copying My Documents folder and using Acronis for "surgical" backups of email and other special data.But right now, all my important data and applications function (for now).

 

[Cerb]

HD Tune gets this data Power On Hours Count: 14,110. Which doesn't seem right from a 2/28/09 purchase. Normal operating temp is 89 or less. Been 89 since I have been examining it - but that is with a wide oen Antec 300 case, and no good airflow.

14110 is on the high side for a desktop, but what makes you think that's not right? It wouldn't be that high with a typical duty cycle, and aggressive power saving, but that's often out of the norm for users here.

Always being 89 just means you're not getting a real reading.

 

[jjsbasmt]

Thanks for expanding my KB. I would think in the interim to a new HD, might not be a bad idea to leave the thing running, turn off write caching in case you get one of those unexpected reboots, and even for gee wow, replace the cable if you have a spare and you could try another sata port on the MB but I doubt that would have any affect.

 

[wjgollatz]

Thanks - I have replaced the cables and checked other ports. I had spare SATA cables stored int he case, and spare ports.

I have a new hard drive now, but will have to reinstall windows. I can't even get an error check to work properly.

 

[wjgollatz]

what would be causing the reboots? Usually, I am trying to copy fiels to another hard drive, or it occurs during a error scan. It happened a few times during an error scan now(never did before).

Right now, I have a 30 day eval instal of Win XP on the space hard drive to run my essentials. My windows copy came from a laptop, that I installed onto a personal built pc, and window will not confirm a new installation onto the new harddrive.

 

[taltamir]

Windows freaks out when it loses connection to a drive. causing a blue screen, except blue screens gave it a bad rep so instead it often does a memory dump (reboot). Potentially first saving a mini dump via black screen.

So... if its not cables (as it still happens after replacement), I would guess either the HDD itself or the mobo (well, the HDD controller which is part of the mobo).

 

[jjsbasmt]

I apologize to all for not suggesting the following sooner, as this is one of the first things I look for when having similar problems...........Is to look the motherboard over very carefully for any bulging or leaking capacitors. Iv'e seen at least a dozen of various brands over the years in fact just had one brought to me in March.